hpvm cloning 11.23 into 11.31 (host) not working boot failure

[antonio elder prado]

Hi,
i have a host hpvm that runs hpux 11.31.
i am doing a dd from the lvol where another machine is located (this is a hpux 11.23). i did create the new vm, added the disks (after dd-ing in the logical volumes copied from previous steps). When i try to boot hpvmconsole into this new hpvm (the 11.23), it does not find any logical volumes, and of course cannot boot, i did try to boot -lm but with not luck.
Can i do a dd and migrate the lgical into another hpvm? is there anything that i should patch for this procedure to work from 11.23 into a 11.31 Host?
in summary this is what i did:
1. went into a hpux machine - 11.23 and dd the root disk
2. created a hpvm on a hpux 11.31 host
3. copied the dd (from #1 into alogical volume on the new host) and added to the new logical volume to the vm with hpvmmodify.
4. hpvmconsole -P newvm
5. on boot it doesn't find any logical voluemes. = panics and reboots.

PS.: i was able to do this pseudo-cloning on a hpvm that runs 11.31 into a host with hpux 11.31.

can you please help me on pointing what i am doing wrong ?

thank you,
./antonio/.
== the output of the boot fail is ====


System Console is on the Built-In Serial Interface
LLT INFO V-14-1-10009 LLT Protocol available
PERFMON: Unrecognized Itanium-? processor model [0x20, 0x1]!
WARNING: Open failed on device 0x1f040002 (error 0x5).
WARNING: Can not read LVM BOOT information (lvmrec).
WARNING: Open failed on device 0x1f040002 (error 0x5).
WARNING: Can not read LVM BOOT information (lvmrec).
WARNING: Open failed on device 0x1f040002 (error 0x5).
WARNING: Can not read LVM BOOT information (BDRA).
WARNING: SWAP device 0xfffffffe is a non-LVM partition, disallowed on LVM disk.
WARNING: SWAP device 0xfffffffe has been deconfigured (set to 0xffffffff).
Swap device table: (start & size given in 512-byte blocks)

System Panic:

panic: swapconf - no primary swap space configured
Stack Trace:
IP Function Name
0xe00000000153a580 adjust_swap_pri+0x840
0xe00000000153ea40 swaptabinit+0x80
0xe000000001536a10 swapinit_phasetwo+0x90
0xe0000000014fa670 DoCalllist+0x3a0
End of Stack Trace

linkstamp: Fri Aug 03 17:07:10 CDT 2007
_release_version: @(#) $Revision: vmunix: B11.23_LR FLAVOR=perf Fri Aug 29 22:35:38 PDT 2003 $

sync'ing disks (0 buffers to flush): (0 buffers to flush):
0 buffers not flushed
0 buffers still dirty

*** A system crash has occurred. (See the above messages for details.)
*** The system is now preparing to dump physical memory to disk, for use
*** in debugging the crash.

ERROR: Your system crashed before I/O and dump configuration was complete.
This system does not support a crash dump under these circumstances.


=========

[melvyn burnard · Accepted solution]

Well the first point to make is that your method is unsupported, so it may or may not work, as you have seen.

The question I ask is whether the 11.23 OS instance is a physical machine, or another Virtual Machine on the same host.
If it is a Virtual Machine on the same host, present another backing store at least the same size, and then on the Guest use DRD to clone it, remove the backing store from this Guest and pressent it to your new VM, reembering that you will have to modify the hostname, ip addtresses etc.

If it is a physical server, then use something like Ignite make_recovery image.

[Eric SAUBIGNAC]

Bonjour,

Don't know why you want to use dd instead of ignite or DRD, as mentionned by Melvyn, but I guess it should work, whatever the source is : a virtual or a physical box. With a physical box as a soure it is a bit more complicated because you will have some maintenance tasks, for example to manage new hardware pathes, but it can work.

Anyway, one possible cause could be the size of the target disk : the size bust be AT LEAST the same size than the source one. Melvin has already underlined it, but I think it must be underlined again ;-) Check this url to see what happens when you reduce the size of the root disk by only a few Mo : http://forums.itrc.hp.com/service/forums/questionanswer.do?threadId=1392503

So, what about 'diskinfo' on source and target disk ?

Eric

[antonio elder prado]

hi Eric and Melvyn,
the hpux 11.23 is a vm, and i want to port into another host (which is 11.31).
the strange thing is that with 11.31 there was no problem on dd-ing from the lvol, and dd-to the new logical volume on the other host, with the 11.31, i just hpvmstart -P newvm and all was fine; however for the 11.23 this method did not work.
actually i never used drd as mentioned by both of you, i will get educated on this.
actually, i did a dd and tried to copy into another machine and it did not work, now i am trying to dd backu into the original host and create another hpvm 11.23 (with the same image as before) but fails in the same manner.
the new fail is something like:
oot device's HP-UX HW path is: 0/0/0/0.0.0
iether0: INITIALIZING HP PCI/PCI-X 1000Base-T at hardware path 0/0/1/0

System Console is on the Built-In Serial Interface
WARNING: Open failed on device 0x1f000002 (error 0x5).
WARNING: Can not read LVM BOOT information (lvmrec).
WARNING: Open failed on device 0x1f000002 (error 0x5).
WARNING: Can not read LVM BOOT information (lvmrec).
WARNING: Open failed on device 0x1f000002 (error 0x5).
WARNING: Can not read LVM BOOT information (BDRA).
WARNING: SWAP device 0xfffffffe is a non-LVM partition, disallowed on LVM disk.
WARNING: SWAP device 0xfffffffe has been deconfigured (set to 0xffffffff).
WARNING: Logical volume for Dump expected but not found.
Swap device table: (start & size given in 512-byte blocks)
WARNING: no swap device configured, so dump cannot be defaulted to primary swap.
WARNING: No dump devices are configured. Dump is disabled.
mountfs:opend

System Panic:

panic: all VFS_MOUNTROOTs failed: NEED DRIVERS ?????
Stack Trace:
IP Function Name
0xe000000000ea3420 vfs_mountroot+0x1d0
0xe0000000011583a0 im_mountroot+0x60
0xe0000000015d8f70 DoCalllist+0x3a0
End of Stack Trace

linkstamp: Fri May 28 14:11:43 EDT 2010
_release_version: @(#) $Revision: vmunix: B11.23_LR FLAVOR=perf Fri Aug 29 22:35:38 PDT 2003 $

sync'ing disks (0 buffers to flush): (0 buffers to flush):

[antonio elder prado]

hi, just to clarify what i want to do

HOST1 HOST 2
hpux 11.31 hpux 11.31
rx2660 rx2660
========== ================
hpvm01 = 11.31 hpvm11 = 11.31 (ok)
hpvm02 = 11.31 hpvm21 = 11.31 (ok)
hpvm03 = 11.23 hpvm31 = 11.23(bad)
hpvm04 = 11.23 hpvm41 = 11.23(bad)

==============================================
method:
on the good and bad, i did as
dd if=/dev/vgvm/lvol01 of=/home/vms/cp01 bs=512k
and this for all 3 of them.
i did then dd back into the other host (host2) and hpvmmodify and tried to start the hpvm.


the size of the logical volumes on host 2 are at least 10G bigger than the original size on the host 1.

thank you guys for your time and help,
./antonio/.

[Eric SAUBIGNAC]

Bonjour Antonio,

Sorry for the late answer, but it was a bank holiday yesterday.

This thread, the thread I have pointed in my previous post (reduction of a vg00 DD) and this new one http://forums.itrc.hp.com/service/forums/questionanswer.do?threadId=1439402 (growth of a vg00 DD) make me wonder if the size of a root disk can be changed on Itanium like boxes ? Maybe due to disk partition table ?

Don't know, but could you realize a new test with the EXACTLY same size on both source and target. Since you are working with LVM as backing store it should not be a problem, right ?

Eric